Redis Streams Integration
Monitor consumer group lag, stream length, and pending entry list (PEL) across your Redis Streams. Detect stalled consumers and PEL bloat before unacknowledged messages become data loss.
How It Works
Deploy redis-exporter via Helm
Install the TigerOps redis-exporter into your cluster. Configure the Redis connection URI with credentials and specify the stream keys or key patterns to monitor via XINFO STREAM and XPENDING commands.
Configure Remote Write
Point the exporter to your TigerOps remote-write endpoint. Stream lengths, consumer group offsets, PEL sizes, and idle consumer counts begin flowing in within seconds of deployment.
Set PEL & Lag Thresholds
Define per-consumer-group pending entry list (PEL) limits and lag thresholds. TigerOps fires alerts when PEL grows beyond your SLO and identifies which consumers have stalled acknowledgements.
Correlate with Redis Server Health
TigerOps links stream consumer lag with Redis memory pressure, eviction rates, keyspace hit ratios, and replication lag — giving full context when streams fall behind.
What You Get Out of the Box
Consumer Group Lag Tracking
Per-consumer and per-group lag tracking using XINFO GROUPS. TigerOps monitors the delta between the last delivered ID and the stream head to quantify how far behind each consumer group is.
Pending Entry List (PEL) Monitoring
Track PEL size per consumer group and per individual consumer. Large PELs indicate stalled acknowledgements or crashed consumers. Alert before unacknowledged messages hit your maxlen trim boundary.
Stream Length & Trim Tracking
Monitor stream entry counts and MAXLEN trim events. Detect when aggressive trimming is dropping unprocessed messages before consumers can read them.
Idle Consumer Detection
Identify consumers that have gone idle and are no longer processing messages. TigerOps tracks per-consumer idle time from XINFO CONSUMERS and alerts on stalled workers.
Redis Server Health Correlation
Correlate stream consumer lag with Redis server-level metrics — memory usage, eviction rates, keyspace hits/misses, connected clients, and replication offset — from the same exporter.
AI Root Cause Analysis
When consumer lag or PEL size spikes, TigerOps AI correlates Redis memory pressure, OOM events, application error rates, and consumer restart patterns to surface the root cause instantly.
Helm Values for redis-exporter
Deploy the TigerOps Redis exporter to your Kubernetes cluster with these Helm values.
# TigerOps redis-exporter Helm values
# helm repo add tigerops https://charts.atatus.net
# helm install redis-exporter tigerops/redis-exporter -f values.yaml
redisExporter:
# Redis connection
redisAddr: redis://redis-master.redis.svc.cluster.local:6379
passwordSecret:
name: redis-credentials
key: password
# Cluster mode (set true for Redis Cluster)
clusterMode: false
# Streams to monitor (supports glob patterns)
streams:
- "orders:events"
- "payments:*"
- "notifications:queue"
# Databases to scan for stream keys
databases: [0, 1, 2]
# How often to run XINFO GROUPS (expensive on large PELs)
streamScrapeInterval: 30s
remoteWrite:
endpoint: https://ingest.atatus.net/api/v1/write
bearerToken: "${TIGEROPS_API_KEY}"
scrapeInterval: 15s
# Alert thresholds
alerts:
consumerGroupLagWarning: 1000 # entries
consumerGroupLagCritical: 10000
pelSizeWarning: 500 # pending entries
pelSizeCritical: 5000
consumerIdleTimeMinutes: 5Common Questions
Which Redis versions does TigerOps support for Streams monitoring?
TigerOps supports Redis 5.0+ for Streams monitoring, as Redis Streams were introduced in Redis 5.0. This includes Redis 6.x and 7.x, Amazon ElastiCache for Redis, Redis Enterprise, and Upstash Redis.
How does TigerOps compute consumer group lag for Redis Streams?
TigerOps uses the XINFO GROUPS command to retrieve the last-delivered-id for each consumer group and the XLEN command for the total stream length. The lag is the count of entries between the last delivered ID and the stream head, computed server-side to avoid transferring entries.
Can TigerOps monitor multiple streams and Redis databases simultaneously?
Yes. Configure a list of stream key patterns in the exporter values and specify multiple Redis databases (0-15). The exporter will discover and monitor all matching streams across all configured databases and automatically track new streams as they appear.
How do I monitor Redis Streams on Redis Cluster mode?
TigerOps supports Redis Cluster. Configure the exporter with cluster mode enabled and provide the seed nodes. The exporter will discover all slots and monitor streams distributed across cluster nodes, aggregating metrics by stream key regardless of which shard they reside on.
What happens when TigerOps detects a stalled consumer in the PEL?
TigerOps fires a consumer stall alert when a consumer has entries in the PEL that exceed your configured idle time threshold. The alert includes the consumer name, group name, stream key, PEL entry count, and the oldest unacknowledged entry ID to aid fast recovery.
Stop Discovering PEL Bloat After Messages Are Lost
Consumer group lag alerts, PEL monitoring, and stalled consumer detection. Deploy in 5 minutes.